☰

Akira Science

From benchtop polymer to bedside implant

Industry
Medical Devices — Bioresorbable Implants
Headquarters
Stockholm, Sweden
Public information as of
February 2026

A4BEE prepared this analysis from publicly available sources. It reflects our own reading of Akira Science's published strategy and is not endorsed by, or produced in cooperation with, Akira Science. Company website

Strategic priorities

Akira Science is a KTH Royal Institute of Technology spin-off developing AkiMed™, a family of programmable bioresorbable copolymers for soft tissue repair. The lead indication is personalised breast reconstruction, with orthopedic and oncology applications in the pipeline. Co-founder and CEO Álvaro Morales López was named to the Forbes 30 Under 30 manufacturing list in 2024, and disclosed funding exceeds four million US dollars across seed, accelerator and grant rounds from 2021 to 2024.

The company describes its long-term model as the 'Manufacturing Hospital': clinics equipped with bioprinters, calibrated storage and validated workflows, producing personalised implants at the point of care rather than shipping from a single factory. Akira's centre of gravity today remains the Stockholm R&D and pilot line, and the technology for scaling production into multiple clinical sites is the work in front of the team.

Implants degrade over seven to eighteen months while new tissue forms, which makes post-market clinical follow-up a long-running data collection exercise rather than a one-time submission. The clinical translation is governed by EU Medical Device Regulation 2017/745 for Class III implantable devices, with Unique Device Identification (UDI) and EUDAMED documentation required for every implant produced.

Two strands of the technology are converging on the same data problem. Computational biomechanics work uses Finite Element Analysis to predict how a scaffold will behave under load. The actual 3D-printing run produces its own telemetry from thermal sensors, filament diameter monitors and motor torque readings. The link between these two data streams is what turns a printed implant into a characterised implant.

Challenges we see

  • Operations Manufacturing

    Keeping polymer storage and printing environments consistent across clinical sites

    AkiMed polymers are sensitive to ambient temperature and humidity. Published data on polycaprolactone and similar poly(ether ester) copolymers shows surface roughness can rise by about 90 percent after 28 days of elevated-temperature storage, altering the mechanical stiffness of the printed scaffold.

    Where the material sits in hospital storage and on the print bed becomes a process variable, and the way to make that variable repeatable across sites is to monitor it rather than to ask each site to remember.

  • Digital Integration

    Linking scaffold simulation to actual printing behaviour

    Akira uses Finite Element Analysis (FEA) to predict scaffold stiffness from a patient scan, and Fused Filament Fabrication (FFF) printers to produce the part. Today the FEA model predicts how the polymer should behave, while the printer produces telemetry on thermal history, filament diameter and motor torque that is not yet fed back into the simulation.

    Closing the loop between simulated scaffold and printed scaffold turns the simulation from a one-shot design check into a model that learns from each batch, which shortens the iteration cycle from bench to bedside.

  • Compliance Regulatory

    Producing a digital record for every implant at every site

    EU MDR 2017/745 requires a Unique Device Identifier and an EUDAMED entry for every Class III implant, linking the device to the filament batch, the printer serial number, the calibration state at the time of print, and the receiving patient.

    Generating that record at the printer rather than compiling it afterwards is the difference between an audit trail that holds up and one that has to be reconstructed, especially when production is distributed across multiple hospitals.

  • Digital Integration

    Protecting patient-specific implant blueprints across hospital networks

    Each AkiMed implant is derived from a patient's CT or MRI scan and constitutes protected health information (PHI). The 'Manufacturing Hospital' model puts commercial 3D printers and patient imaging data on the same hospital network as general clinical systems.

    Once the printer sits on a hospital network, the security boundary follows the patient blueprint through the printer, the storage cabinet and the maintenance laptop, so security has to be designed at the same time as the network.

  • Operations Manufacturing

    Detecting scaffold defects without an expert at the printer

    FFF printing is sensitive to nozzle clogs, melt viscosity shifts and motor torque variations, all of which can introduce defects into the scaffold mesh. The Manufacturing Hospital model is designed to run without a polymer or bioprinting specialist on site.

    Layer-by-layer monitoring has to do what an experienced operator would do today, with the difference between a caught defect and a shipped defect visible at inspection rather than during surgery.

Opportunities, by urgency and business impact

Each bubble is one opportunity, numbered to match the list below. Further right means it bites sooner; higher means a bigger effect on the business. A bigger bubble means a bigger implementation effort.

Source: A4BEE analysis of public sources

  1. Monitoring polymer storage and print conditions at every Manufacturing Hospital

    AkiMed's mechanical properties shift with storage temperature, humidity and time. With implants being produced at multiple clinical sites, a manual approach to environmental control leaves each batch subject to the conditions of the room it happens to be in.

    Instrumented storage cabinets and print enclosures stream temperature, humidity and CO2 readings into a shared view, so the environmental history of each filament batch and each scaffold is recorded alongside the implant record.

    • Impact of storage at different thermal conditions on surface characteristics of 3D printed polycaprolactone and poly(ε-caprolactone-co-p-dioxanone) scaffolds, ResearchGate, 2023
    • Akira Science AkiMed product information

  2. Connecting FEA simulation and printer telemetry through a shared data layer

    Finite Element Analysis predicts how a scaffold should behave under load, but the actual print produces its own data on thermal history, filament diameter and motor torque. The two streams live in separate systems today, so each printed scaffold is interpreted against an idealised simulation rather than against what the printer actually did.

    A shared data layer that ingests printer telemetry and aligns it to the FEA model turns each print into a calibration point, and over time the simulation reflects the behaviour of Akira's materials in Akira's printers rather than a textbook material model.

    • Akira Science Beyond Breast programme description
    • Industry 4.0 and the Future of the Pharmaceutical Industry, ResearchGate, 2021

  3. Generating UDI, EUDAMED and PMCF records at the printer

    EU MDR 2017/745 asks for a Unique Device Identifier, an EUDAMED entry and a Post-Market Clinical Follow-Up plan for every implant. With production distributed across hospitals, producing these records by hand for each implant creates the conditions for transcription errors and audit findings.

    A device record assembled at the printer captures filament batch, printer serial number, calibration state and patient reference automatically, and the same record structure supports the periodic PMCF data collection that runs over the seven-to-eighteen-month degradation window.

    • Expanding Quality by Design Principles to Support 3D Printed Medical Device Development Following the Renewed Regulatory Framework in Europe, MDPI, 2022
    • 3D Printing Facility in Healthcare Institutions: An Effective Strategy to be Compliant with Medical Device Regulation EU 2017/745, Crimson Publishers

  4. Designing IT/OT security for distributed implant production

    Patient-specific implant blueprints and the printers that produce them now sit inside hospital networks alongside general clinical systems. Commercial 3D printers are not engineered for that environment, and the data they handle is protected health information.

    A security design that segments the printer and the blueprint vault from the general hospital network, applies identity checks to every device and user, and aligns to IEC 62443 lets each Manufacturing Hospital operate inside a defined zone rather than inside an open network.

    • Cybersecurity vulnerabilities in medical devices: a complex environment and multifaceted problem, PMC, 2015
    • Securing Operational Technologies used for medical product manufacturing, GE Healthcare, 2025

  5. Layer-by-layer inspection of the printed scaffold

    FFF printing produces defects from nozzle clogs, viscosity shifts and torque variations. Without an experienced polymer scientist at the printer, those defects reach the scaffold and then the patient.

    Computer vision monitoring on the print bed, combined with rules that flag or pause on detected anomalies, gives a printed scaffold an inspection record that is generated at production time and that flags a layer problem before the next layer is laid down.

    • Influence of Ambient Temperature and Crystalline Structure on Fracture Toughness and Production of Thermoplastic by Enclosure FDM 3D Printer, ResearchGate, 2023
    • Challenges and Opportunities in 3D Printing of Biodegradable Medical Devices by Emerging Photopolymerization Techniques, ETH Zurich, 2024

What we'd propose

  • Digital CDMO

    Environmental monitoring and control for the Manufacturing Hospital

    We deploy an IoT-based environmental monitoring and control layer for the polymer storage cabinets and print enclosures at each Manufacturing Hospital, streaming temperature, humidity and CO2 into a shared view alongside the implant record.

    • Environmental sensor network

      Sensors at storage and at the printer

      Calibrated temperature, humidity and CO2 sensors connect to a shared time-series store over OPC UA or MQTT, so the environmental history of each filament batch and each print is captured rather than assumed.

    • Adaptive climate control

      Storage that holds its setpoint

      Climate control in storage and print enclosures reacts to the sensor stream rather than to a fixed schedule, with setpoints defined per material grade so an AkiMed batch and a research-grade polymer can share a room without sharing a tolerance.

    • Compliance dashboard

      The conditions a batch actually saw

      A dashboard surfaces environmental conditions, setpoint excursions and per-batch history in a form that supports the device record and the EUDAMED entry, so an inspector can read the conditions the polymer was stored under rather than a manual log.

    • Each filament batch carries a record of the conditions it saw between synthesis and print.
    • Environmental setpoints are defined per material rather than per room.
    • The same data stream serves operations, quality and regulatory reporting.

    We've done this before

    See how we do Digital CDMO
  • Enterprise AI

    Shared data layer for FEA simulation and printer telemetry

    We build a shared data layer that ingests printer telemetry, aligns it to the FEA model and serves the combined view to design, manufacturing and clinical teams, so each print updates the simulation rather than ending at it.

    • Printer telemetry ingestion

      Data off every FFF run

      Thermal, filament diameter and motor torque sensors on the printer stream into a time-series store, with each reading tagged to the print job, the material grade and the printer serial number.

    • Simulation feedback loop

      FEA informed by what the printer did

      The FEA model is parameterised against the actual print history, so the predicted stiffness of a new scaffold reflects Akira's materials in Akira's printers rather than a textbook material.

    • Parameter recommendations

      Settings derived from history

      On each new design, recommended print speed, nozzle temperature and fan settings are drawn from similar past prints and surfaced to the operator as a starting point, with the next print updating the recommendation.

    • Each print becomes a calibration point for the next design, rather than an isolated run.
    • Design reviews compare predicted and measured behaviour on the same scaffold, not on two different artefacts.
    • The shared data layer is reusable as Akira adds printer types and material grades.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Automated device record for UDI, EUDAMED and PMCF

    We connect the printer, the filament batch record and the patient reference so every implant carries a structured device record from the moment it is printed, and the same record structure supports EUDAMED registration and Post-Market Clinical Follow-Up data collection.

    • Automated UDI generation

      A device identifier at the printer

      A Unique Device Identifier is generated at the printer for each implant, linking the device to the filament batch, the printer serial number and the calibration state at the time of print, with the identifier written to the EUDAMED-compatible record.

    • End-to-end traceability

      From filament to patient

      A digital thread connects the filament batch certificate, the printer telemetry, the operator's confirmation and the patient reference, so an inspector can follow one implant back to its source material in a single view.

    • PMCF data collection

      Long-window follow-up as data

      The same record structure carries the implant through Post-Market Clinical Follow-Up, capturing the planned follow-up visits and the data points that need to be collected over the seven-to-eighteen-month degradation window.

    • The device record is produced by the line that makes the device, not assembled after it.
    • EUDAMED submission and PMCF data collection share a single source of truth.
    • An audit question about an implant is answered from the record rather than reconstructed.

    We've done this before

    See how we do Digital Lab
  • Digital CDMO

    IT and OT security design for Manufacturing Hospital nodes

    We define a security architecture for each Manufacturing Hospital that segments the printer and the blueprint vault from the general hospital network, applies identity checks to every device and user, and aligns to IEC 62443 for the operational technology in scope.

    • Network segmentation

      Printer and vault on a defined zone

      The printer, the blueprint vault and the maintenance workstation sit on a network segment that is isolated from the hospital's general clinical network, with controlled conduits for the data that has to cross the boundary.

    • Identity-based access

      Every device and user verified

      Every device and user that touches the printer, the blueprint vault or the device record is authenticated and authorised, with identity checks replacing the assumption that anything on the network segment is trusted.

    • IEC 62443 baseline

      OT security aligned to a published standard

      The security architecture follows IEC 62443 for the operational technology in scope, with the zones, conduits and access rules documented in the form expected by an EU MDR notified body reviewing a Class III device process.

    • The patient blueprint is held inside a defined security zone rather than on the open hospital network.
    • Every action on a printer or a vault entry is attributable to an authenticated identity.
    • The IEC 62443 alignment gives the notified body a documented baseline to review.

    We've done this before

    See how we do Digital CDMO
  • Digital Lab

    Computer vision monitoring of the printing process

    We deploy camera-based layer monitoring and rules-based anomaly detection on the FFF printer so each scaffold is inspected as it is printed and a layer problem is flagged before the next layer is laid down.

    • Layer-by-layer monitoring

      A camera on every layer

      An imaging system on the print enclosure captures each layer as it is deposited, with frame timing aligned to the printer's job clock so a defect can be tied back to the layer number and the print parameters in effect.

    • Anomaly detection and pause

      Flag or pause on a defect

      Rules trained on Akira's defect catalogue flag nozzle clogging, layer adhesion failure and mesh irregularities, with a configurable response that records the anomaly and pauses the print for operator review on critical defects.

    • Inspection record per print

      The scaffold's inspection history

      Each print ends with an inspection record carrying the layer images, the anomaly log and the final disposition, which attaches to the device record and to the EUDAMED entry rather than sitting in a separate quality system.

    • The scaffold is inspected at production time, not at surgery.
    • An anomaly is tied to a specific layer and a specific print run, so an investigation starts from a record.
    • The inspection history travels with the device record into the EUDAMED entry.

    We've done this before

    See how we do Digital Lab

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what Akira Science's own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Data integration 30 → 85
Patient scan, Finite Element Analysis model and printer telemetry sit in separate systems today. Bringing them into a shared layer is what makes a printed scaffold a characterised scaffold.
Process automation 25 → 78
Environmental monitoring, anomaly detection and UDI generation are largely manual in the current Stockholm line. The Manufacturing Hospital model asks them to be automatic at every clinical site.
Regulatory compliance 38 → 92
EU MDR 2017/745 asks for a UDI, an EUDAMED entry and a PMCF plan for every Class III implant. The record today is assembled after the implant is produced, which is the gap the automation closes.
Cybersecurity 22 → 85
Patient blueprints travel from the imaging system into the printer, and the printer sits on a hospital network. The current setup assumes a trusted network rather than designing for one that is not.
Scalability 32 → 82
Akira describes the Manufacturing Hospital model as a distributed network of clinical sites. Replicating the Stockholm line at every site requires the same reference architecture, the same security baseline and the same data view at each site.
Digital twin maturity 28 → 80
The FEA simulation predicts how a scaffold should behave. The link between predicted and measured behaviour is the step that moves the simulation from a one-shot design check to a model that is updated by every print.

Check this yourself

Our Service Portal has free self-assessments and market comparisons. These are the ones that line up with what we've read above — no sales call required.

Think we've read this right?

Talk to us

Related reading

This is an independent analysis prepared by A4BEE from publicly available information as of February 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with Akira Science, and may be incomplete or inaccurate. All company names and trademarks are the property of their respective owners. To request a correction or removal, contact [email protected].